index.md

layout
default

We present audio examples of our model IAFF-VC. To perform any-to-any voice conversion, we set two scenarios: seen-to-seen (S2S) and unseen-to-unseen(U2U). In each scenario, we present four converted speech: male(M)-male(M), female(F)-female(F), male(M)-female(F), and female(F)-male(M). Furthermore, we perform the conversion in multi-utterance using 3 samples. Same scenarios are set in multi-utterance.

Audio examples

S2S: seen-to-seen; U2U: unseen-to-unseen

F: Female; M:Male

Any-to-Any conversion

S2S:

F-F:

• source:

• target:

• converted:

F-M:

• source:

• target:

• converted:

M-M:

• source：

• target:

• converted:

M-F:

• source：

• target:

• converted:

U2U:

F-F:

• source:

• target:

• converted:

F-M:

• source:

• target:

• converted:

M-M:

• source：

• target:

• converted:

M-F:

• source：

• target:

• converted:

Multi-utterance conversion

S2S:

F-F:

• source:

• target:

• converted:

F-M:

• source:

• target:

• converted:

M-M:

• source：

• target:

• converted:

M-F:

• source：

• target:

• converted:

U2U:

F-F:

• source:

• target:

• converted:

F-M:

• source:

• target:

• converted:

M-M:

• source：

• target:

• converted:

M-F:

• source：

• target:

• converted: